On 12/09/2013 10:01 AM, Richard Biener wrote:
On Mon, Dec 9, 2013 at 3:49 PM, Kenneth Zadeck wrote:
On 12/08/2013 05:35 AM, Richard Biener wrote:
Richard Sandiford wrote:
Kenneth Zadeck writes:
#define WIDE_INT_MAX_ELTS \
- ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1)
On 12/09/2013 10:12 AM, Richard Sandiford wrote:
Richard Biener writes:
On Mon, Dec 9, 2013 at 3:49 PM, Kenneth Zadeck wrote:
On 12/08/2013 05:35 AM, Richard Biener wrote:
Richard Sandiford wrote:
Kenneth Zadeck writes:
#define WIDE_INT_MAX_ELTS \
- ((4 * MAX_BITSIZE_MODE_ANY_INT +
Richard Biener writes:
> On Mon, Dec 9, 2013 at 3:49 PM, Kenneth Zadeck
> wrote:
>> On 12/08/2013 05:35 AM, Richard Biener wrote:
>>>
>>> Richard Sandiford wrote:
Kenneth Zadeck writes:
>
> #define WIDE_INT_MAX_ELTS \
> - ((4 * MAX_BITSIZE_MODE_ANY_INT +
On 12/09/2013 10:01 AM, Richard Biener wrote:
On Mon, Dec 9, 2013 at 3:49 PM, Kenneth Zadeck wrote:
On 12/08/2013 05:35 AM, Richard Biener wrote:
Richard Sandiford wrote:
Kenneth Zadeck writes:
#define WIDE_INT_MAX_ELTS \
- ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1)
On Mon, Dec 9, 2013 at 3:49 PM, Kenneth Zadeck wrote:
> On 12/08/2013 05:35 AM, Richard Biener wrote:
>>
>> Richard Sandiford wrote:
>>>
>>> Kenneth Zadeck writes:
#define WIDE_INT_MAX_ELTS \
- ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
>>>
On 12/08/2013 05:35 AM, Richard Biener wrote:
Richard Sandiford wrote:
Kenneth Zadeck writes:
#define WIDE_INT_MAX_ELTS \
- ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
- / HOST_BITS_PER_WIDE_INT)
+ (((MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1)\
+
Richard Sandiford wrote:
>Kenneth Zadeck writes:
>>#define WIDE_INT_MAX_ELTS \
>> - ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
>> - / HOST_BITS_PER_WIDE_INT)
>> + (((MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
>> +/ HOST_BITS_PE
Kenneth Zadeck writes:
>#define WIDE_INT_MAX_ELTS \
> - ((4 * MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
> - / HOST_BITS_PER_WIDE_INT)
> + (((MAX_BITSIZE_MODE_ANY_INT + HOST_BITS_PER_WIDE_INT - 1) \
> +/ HOST_BITS_PER_WIDE_INT) + 1)
I think t
On 12/06/2013 01:32 PM, Richard Sandiford wrote:
Kenneth Zadeck writes:
On 12/03/2013 11:52 AM, Richard Sandiford wrote:
Kenneth Zadeck writes:
Index: tree-vrp.c
===
--- tree-vrp.c (revision 205597)
+++ tree-vrp.c (working cop
Kenneth Zadeck writes:
> On 12/03/2013 11:52 AM, Richard Sandiford wrote:
>> Kenneth Zadeck writes:
>>> Index: tree-vrp.c
>>> ===
>>> --- tree-vrp.c (revision 205597)
>>> +++ tree-vrp.c (working copy)
>>> @@ -2611,22 +2611,
Richard asked to see the patch where i did the changes the way that he
asked in his email. Doing it the way he wants potentially has advantages
over the way that i did it, but his technique fails for non obvious
reasons. The failure in building is:
g++ -c -g -DIN_GCC -fno-exceptions -fno-rtti
On 12/03/2013 11:52 AM, Richard Sandiford wrote:
Kenneth Zadeck writes:
Index: tree-vrp.c
===
--- tree-vrp.c (revision 205597)
+++ tree-vrp.c (working copy)
@@ -2611,22 +2611,28 @@ extract_range_from_binary_expr_1 (value_
On Dec 3, 2013, at 8:52 AM, Richard Sandiford
wrote:
> You probably already realise this, but for avoidance of doubt, Richard
> was also asking that we reduce MAX_BITSIZE_MODE_ANY_INT to 128 on x86_64,
> since that's the largest scalar_mode_supported_p mode.
Personally, I'd love to see scalar_mo
Kenneth Zadeck writes:
> Index: tree-vrp.c
> ===
> --- tree-vrp.c(revision 205597)
> +++ tree-vrp.c(working copy)
> @@ -2611,22 +2611,28 @@ extract_range_from_binary_expr_1 (value_
>
>signop sign = TYPE_SIGN
On 11/29/2013 05:24 AM, Richard Biener wrote:
On Thu, Nov 28, 2013 at 6:11 PM, Kenneth Zadeck
wrote:
This patch does three things in wide-int:
1) it cleans up some comments.
2) removes a small amount of trash.
3) it changes the max size of the wide int from being 4x of
MAX_BITSIZE_MODE_ANY_INT
committed as revision 205599 to wide-int branch.
kenny
On 12/02/2013 05:50 AM, Richard Biener wrote:
On Sat, Nov 30, 2013 at 1:55 AM, Kenneth Zadeck
wrote:
Richi,
this is the first of either 2 or 3 patches to fix this.There are two
places that need be fixed for us to do 1X + 1 and this p
On Sat, Nov 30, 2013 at 1:55 AM, Kenneth Zadeck
wrote:
> Richi,
>
> this is the first of either 2 or 3 patches to fix this.There are two
> places that need be fixed for us to do 1X + 1 and this patch fixes the first
> one. There was an unnecessary call to mul_full and this was the only call
Richi,
this is the first of either 2 or 3 patches to fix this.There are two
places that need be fixed for us to do 1X + 1 and this patch fixes the
first one. There was an unnecessary call to mul_full and this was the
only call to mul_full. So this patch removes the call and also the
f
could be.
On 11/29/2013 06:57 AM, Richard Sandiford wrote:
Looks good to me FWIW, except:
Kenneth Zadeck writes:
@@ -112,11 +114,11 @@ along with GCC; see the file COPYING3.
two, the default is the prefered representation.
All three flavors of wide_int are represented as a vector
On Nov 29, 2013, at 4:24 AM, Richard Biener wrote:
> On Thu, Nov 28, 2013 at 6:11 PM, Kenneth Zadeck
> wrote:
>> This patch does three things in wide-int:
>>
>> 1) it cleans up some comments.
>> 2) removes a small amount of trash.
>> 3) it changes the max size of the wide int from being 4x
Looks good to me FWIW, except:
Kenneth Zadeck writes:
> @@ -112,11 +114,11 @@ along with GCC; see the file COPYING3.
> two, the default is the prefered representation.
>
> All three flavors of wide_int are represented as a vector of
> - HOST_WIDE_INTs. The default and widest_int vect
On Fri, Nov 29, 2013 at 12:26 PM, Richard Sandiford
wrote:
> Richard Biener writes:
>> On Thu, Nov 28, 2013 at 6:11 PM, Kenneth Zadeck
>> wrote:
>>> This patch does three things in wide-int:
>>>
>>> 1) it cleans up some comments.
>>> 2) removes a small amount of trash.
>>> 3) it changes the max
Richard Biener writes:
> On Thu, Nov 28, 2013 at 6:11 PM, Kenneth Zadeck
> wrote:
>> This patch does three things in wide-int:
>>
>> 1) it cleans up some comments.
>> 2) removes a small amount of trash.
>> 3) it changes the max size of the wide int from being 4x of
>> MAX_BITSIZE_MODE_ANY_INT to
On Thu, Nov 28, 2013 at 6:11 PM, Kenneth Zadeck
wrote:
> This patch does three things in wide-int:
>
> 1) it cleans up some comments.
> 2) removes a small amount of trash.
> 3) it changes the max size of the wide int from being 4x of
> MAX_BITSIZE_MODE_ANY_INT to 2x +1. This should improve large
This patch does three things in wide-int:
1) it cleans up some comments.
2) removes a small amount of trash.
3) it changes the max size of the wide int from being 4x of
MAX_BITSIZE_MODE_ANY_INT to 2x +1. This should improve large muls and
divs as well as perhaps help with some cache behavior.
25 matches
Mail list logo